Abstract :
Anorexia nervosa (AN) is a chronic eating disorder in which body image distortion, chronically low body weight, excessive dieting, purging and amenorrhea dominate the clinical picture. This disorder affects 0.5% of women. Family studies reveal that 5% of first degree relatives of AN probands will have AN or bulimia. Monozygotic twin concordance rates ( 55%) are significantly higher than dizygotic twin concordance rates ( 5%), giving rise to heritability estimates of 50%. Thus, half the variance in susceptibility to AN is explained by additive genetic factors. These estimates for heritability are not very different from those for schizophrenia, bipolar illness and diabetes mellitus, in which molecular linkage studies of affected sibling pairs have been successful in localizing susceptibility genes. While lod score methods may be more powerful than affected sibling pair approaches in a search for loci of major effects, lod score methods are optimally applied when multiplex, extended pedigrees are available. Large multiplex families with many cases of AN have not been reported in the literature. Since such extended AN kindreds are rare, the affected sibling pair method offers an invaluable alternative. For a sample of 200 affected sibling pairs, this method provide excellent power to detect linkage to a locus which increases risk for AN by a factor of two or more. Methods for a genomewide search for AN susceptibility loci will be described.